Storage Structure and Hierarchy

Figure 1.8: Going down the hierarchy

• Decreasing cost per bit
• Increasing capacity
• Increasing access time
• Decreasing frequency of access of the memory by the processor
• Fully electronic memory systems are the fastest and most expensive, hence must be used in cost effective ways. This memory system is called main memory or primary memory.
• A source of cheaper-per-byte and non-volatile storage is provided by magnetic disk. However, the computer does not have direct random access to any byte at any time on the disk - the magnetic discs in the drive are rotating and magnetic heads move in and out in order to access any part of the surface area on the disc that holds data. This means access usually involves a disc rotation delay and also a head positioning delay.
• Other common forms of non-volatile secondary storage include: optical CD drives (CD-R write-once or CD-RW read-write), recent flash memory chips in very small modules that can be inserted into laptop card interfaces or can be used for data logging.
• Stages such as the CPU registers and cache are typically located within the CPU chip so distances are very short and busses can be made very very wide (e.g. 128-bits), yielding very fast speeds.
• Future storage technology includes 3-dimensional crystal structures which allow optical access to a dense 3-dimensional storage facility.